论文阅读 [TPAMI-2022] End-to-End Optimized Versatile Image Compression With Wavelet-Like Transform
论文搜索(studyai.com)
搜索论文: End-to-End Optimized Versatile Image Compression With Wavelet-Like Transform
搜索论文: http://www.studyai.com/search/whole-site/?q=End-to-End+Optimized+Versatile+Image+Compression+With+Wavelet-Like+Transform
关键字(Keywords)
Image coding; Quantization (signal); Wavelet transforms; Bit rate; Entropy coding; Rate-distortion; Deep network; end-to-end optimization; image compression; lossless compression; lossy compression; wavelet transform
机器学习; 机器视觉
自编码器; 图像压缩
摘要(Abstract)
Built on deep networks, end-to-end optimized image compression has made impressive progress in the past few years.
基于深度网络的端到端优化图像压缩在过去几年中取得了令人瞩目的进展。.
Previous studies usually adopt a compressive auto-encoder, where the encoder part first converts image into latent features, and then quantizes the features before encoding them into bits.
以往的研究通常采用压缩式自动编码器,其中编码器部分首先将图像转换为潜在特征,然后对特征进行量化,然后再将其编码为比特。.
Both the conversion and the quantization incur information loss, resulting in a difficulty to optimally achieve arbitrary compression ratio.
转换和量化都会导致信息丢失,导致难以以最佳方式实现任意压缩比。.
We propose iWave++ as a new end-to-end optimized image compression scheme, in which iWave, a trained wavelet-like transform, converts images into coefficients without any information loss.
我们提出iWave++作为一种新的端到端优化图像压缩方案,在这种方案中,iWave是一种经过训练的类小波变换,可以在不丢失任何信息的情况下将图像转换为系数。.
Then the coefficients are optionally quantized and encoded into bits.
然后可选地对系数进行量化并编码为比特。.
Different from the previous schemes, iWave++ is versatile: a single model supports both lossless and lossy compression, and also achieves arbitrary compression ratio by simply adjusting the quantization scale.
与之前的方案不同,iWave++是多功能的:单个模型支持无损和有损压缩,还可以通过简单地调整量化比例来实现任意压缩比。.
iWave++ also features a carefully designed entropy coding engine to encode the coefficients progressively, and a de-quantization module for lossy compression.
iWave++还具有一个精心设计的熵编码引擎,用于逐步编码系数,以及一个用于有损压缩的去量化模块。.
Experimental results show that lossy iWave++ achieves state-of-the-art compression efficiency compared with deep network-based methods; on the Kodak dataset, lossy iWave++ leads to 17.34 percent bits saving over BPG; lossless iWave++ achieves comparable or better performance than FLIF.
实验结果表明,与基于深度网络的方法相比,有损iWave++实现了最先进的压缩效率;在柯达数据集上,有损iWave++比BPG节省了17.34%的比特数;无损iWave++实现了与FLIF相当或更好的性能。.
Our code and models are available at https://github.com/mahaichuan/Versatile-Image-Compression…
我们的代码和模型可在https://github.com/mahaichuan/Versatile-Image-Compression…
作者(Authors)
[‘Haichuan Ma’, ‘Dong Liu’, ‘Ning Yan’, ‘Houqiang Li’, ‘Feng Wu’]